Emergent Dynamics of Benefit and Cost of Learning in the Course of Evolution

An interaction between evolution and learning called the Baldwin effect is known as the two-step evolution caused by the balances between benefit and cost of learning. However, little is still known about dynamic properties of these balances in complex environments. Our purpose is to clarify the key factors and conditions which bring about the dynamic evolution of the benefit and cost of learning by focusing on the quantitative evolution of phenotypic plasticity under the assumption of epistatic interactions. We have constructed an evolutionary model of quantitative traits by using an extended version of NK fitness landscape, in which plasticity of each phenotype is genetically defined and plastic phenotypes can be adjusted by learning iterations. The simulation results have shown that the dynamic emergence of the benefit and cost of learning brought about the threestep evolution of the Baldwin effect (observed in our previous study [8]) and it depends on the balances between the degree of epistasis and the number of learning iterations.